Upload
others
View
7
Download
0
Embed Size (px)
Citation preview
Confidential | Copyright © 2017 IHS Markit Ltd
An introduction to the PMI
surveys July 18, 2017
An introduction to the PMI surveys
| 2
Introduction 3
Methodology 4
What IHS Markit do and why we do it 4
Creating IHS Markit PMI Data 5
Panel building 6
Data Collection 6
Data Calculation 7
Seasonal adjustment 7
Coverage: Nation 8
Coverage: Sector 9
Coverage: Survey Indices 11
Coverage: Survey Indices –
Headline Indices 13
Coverage: Survey Indices –
Flash PMI Indices 14
An introduction to the PMI surveys
| 3
Introduction
Welcome to the IHS Markit PMI™ data, compiled by IHS Markit. The following document provides a simple single source overview of IHS Markit, the IHS Markit PMI survey methodology and the data produced.
IHS Markit is a global provider of some of the world’s most influential business surveys, one example being the PMI.
Our indicators are closely watched by the financial markets and help shape economic policy, by providing some of the earliest signals of economic performance.
The surveys form a key source of business intelligence for participants operating in the ‘real’ economy. Our data are presented in a clear, jargon-free format, allowing a swift understanding of current economic trends.
Purchasing Managers’ Index™ (PMI™) series are monthly economic surveys of carefully selected companies. They provide an advance signal of what is really happening in the private sector economy, by tracking variables such as output, new orders, employment and prices across key sectors.
Before outlining the PMI methodology in greater detail, take a look at the charts to the right showing examples of the close relationship between the PMI survey indicators and official economic data. The PMI data – which are published earlier than the official data – clearly identify the turning points in the business cycle and closely track the rate of change in key economic variables such as gross domestic product (GDP). This proven track record is one of the many reasons why the PMI is so closely monitored.
An introduction to the PMI surveys
| 4
Methodology
What IHS Markit do and why we do it
The PMI methodology was developed to resolve issues caused by the following deficiencies and weaknesses in many official economic indicators:
Inconsistent methodologies between nations and regions, making like-for-like comparisons between nations or regions difficult
Infrequent releases and delays in publication
Difficulties in providing coverage for the whole economy
Frequent and ongoing revisions
These problems mean decisions are often made on the basis of data that subsequently change, providing the potential for policy and investment allocation errors.
The PMI methodology allows for rapid collection, processing, analysis and publication of data, providing a timely, accurate, reliable and internationally consistent series of global economic indicators.
There are four key elements to the methodology:
1. Standard questionnaire asked to key decision makers
Companies asked questions covering output, demand, prices, employment
Providing ‘grass roots’ level perspective on actual economic trends
Designed to produce timely information ahead of official data
2. Responses are provided simply on the basis if levels are higher, the same or lower than in the prior month
Allows rapid and accurate data collection
Quantitative responses ensure hard economic data…based on fact NOT opinions
3. Data are collected from structured panels, recruited and managed to ensure accurate representation of underlying economic structure
Resulting indices are also weighted to further ensure accuracy and representation
4. Underlying survey responses included in the indices are not revised after first publication
An introduction to the PMI surveys
| 5
Creating IHS Markit PMI Data
The methodology for creating IHS Markit PMI data flows through a number of stages, starting from the initial building of the panel, through data collection and calculation and then the final publication of the latest readings. The following diagram outlines the process.
Panel is recruited to accurately
represent the underlying structure of
the sector as given by official GVA
and company size data.
Data are collected on a monthly
basis. Relationships are built with
panel members to ensure low
‘churn’.
Raw ‘diffusion’ indices are
calculated for the survey
sub-indices, Data are
weighted to ensure they are
representative.
Weighted raw diffusion indices are
seasonally adjusted to account for
expected seasonal variations.
An introduction to the PMI surveys
| 6
Panel building
Structural economic indicators obtained from national statistics offices are used to construct a
recruitment matrix by company size and sector contributions to overall sectorial gross value added
(GVA). This matrix is then used to target recruitment of a representative sample of companies for the
survey. The building of the panel may be undertaken either by an in-house panel manager at IHS
Markit or by a country local agency.
An example of a recruitment matrix is shown below:
ISIC Service Sector Classification 1-19 Employees 20-249 Employees 250+ Employees Sector (%)
Hotels & Restaurants 9 7 3 4
Transport & Storage 8 15 18 8
Post & Telecommunications 0 25 0 5
Financial Intermediation 12 14 17 8
Renting & Business Activity 98 108 87 58
Other Services 24 28 34 17
Total 151 197 163 100
*Please note actual targets are more detailed by sector than illustrated
(return to section Data Calculation)
Data Collection
Data are collected either by an in-house panel manager at IHS Markit or by a country local agency.
Companies are asked a standard sector-specific questionnaire on a monthly basis. Questions are of
the form, using output as an example, of:
“Is the level of output at your unit (in volume terms) higher, the same or lower than one month ago”
Companies are also asked to provide reasons (freehand) for the response given.
Data are usually collected during the middle two weeks of the month, so companies are asked to
compare mid-month to mid-month. All responses are input into IHS Markit Polling Engine database
for data calculation and weighting purposes.
Details on the variables for which data are collected are included in section Coverage: Survey Indices
An introduction to the PMI surveys
| 7
Data Calculation
Individual company responses are weighted according to sector and company size contributions to
overall sectorial gross value added based on the weighting matrix constructed in the same way as the
recruitment matrix used when recruiting the panel. The weighting provides IHS Markit with the
weighted proportions of companies responding up, the same and lower.
The latest values for the sub-indices are then calculated using the following formula:
INDEX VALUE = (% ‘up’) + (0.5 * (% ‘the same’)) + (0.0 * (% reporting ‘down’))
The resulting index values are therefore bounded between 0 (all companies respond lower) and 100
(all companies respond higher), with a theoretical no-change mark at 50 (all companies respond ‘the
same’ or equal proportions respond up as do for down).
The resulting index is called a diffusion index, with values above 50.0 signaling expansion and below
50.0 indicating contraction. The distance from the 50 no-change mark signals the implied rate of
change in the variable, the further from 50.0 the faster the rate of change indicated.
Note on interpreting the index
For example, take a movement in a PMI output sub-index from 55.0 in January to 52.5 in February.
Although the level of the output index has fallen, it has nonetheless posted above 50.0 in both
months. The correct interpretation is that the volume of output expanded in both January and
February, but that the rate of increase was slower in February compared to January.
Seasonal adjustment
IHS Markit PMI sub-indices are seasonally adjusted using a combination of the Census Bureau X-12
ARIMA method and an in-house method developed by IHS Markit incorporating past adjustment
factors and additional information on observed seasonality provided from panel members’ responses.
An introduction to the PMI surveys
| 8
Coverage: Nation
Purchasing Managers’ Index (PMI) surveys are now available for more than 30 countries and for key regions including the Eurozone. Together these countries account for over 86% of global gross domestic product (GDP), providing a true, internationally consistent indication of global economic trends based on fact, not opinion. They are the most closely-watched business surveys in the world, favoured by central banks, financial markets and business decision makers for their ability to provide up-to-date, accurate and often unique monthly indicators of economic trends.
IHS Markit publishes PMI data for the following nations (correct at July 18 2017):
Australia
Austria
Brazil
Canada
China
Czech Republic
Egypt
France
Germany
Greece
Hong Kong
India
Indonesia
Ireland
Italy
Japan
Kenya
Lebanon
Malaysia
Myanmar
Mexico
Nigeria
Netherlands
Philippines
Poland
Russia
Saudi Arabia
Singapore
South Africa
South Korea
Spain
Taiwan
Thailand
Turkey
Uganda
United Arab
Emirates
United Kingdom
United States
Vietnam
Zambia
An introduction to the PMI surveys
| 9
Coverage: Sector The following table outlines the manufacturing, services and whole economy surveys produced by
IHS Markit (correct as of July 18 2017):
Nation\Region PMI
Manufacturing
(Monthly)
PMI
Services
(Monthly)
PMI
Whole Economy* or Composite**
(Monthly)
Outlook
Manufacturing
(Tri-annual)
Outlook
Services
(Tri-annual)
Outlook
Composite**
(Tri-annual)
Global Y Y Y** Y Y Y**
European Union
Eurozone
Germany
France
Italy
Spain
Netherlands
Austria
Ireland
Greece
Non-Eurozone EU
United Kingdom
Czech Republic
Poland
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
–
–
Y
–
Y
–
–
Y**
Y**
Y**
Y**
Y**
Y**
–
–
Y**
–
Y**
–
–
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
–
–
Y
–
Y
–
–
Y**
Y**
Y**
Y**
Y**
Y**
–
–
Y**
–
Y**
–
–
Americas
USA
Canada
Mexico
Brazil
Y
Y
Y
Y
Y
–
–
Y
Y**
–
–
Y**
Y
–
–
Y
Y
–
–
Y
Y**
–
–
Y**
Asia
China
Japan
India
South Korea
Indonesia
Taiwan
Vietnam
Malaysia
Myanmar
Thailand
Philippines
Singapore
Hong Kong
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
–
–
Y
Y
Y
–
–
–
–
–
–
–
–
–
–
Y**
Y**
Y**
–
–
–
–
–
–
–
–
Y*
Y*
Y
Y
Y
–
–
–
–
–
–
–
–
–
–
Y
Y
Y
–
–
–
–
–
–
–
–
–
–
Y**
Y**
Y**
–
–
–
–
–
–
–
–
–
–
An introduction to the PMI surveys
| 10
Middle East and Africa
Egypt
Saudi Arabia
UAE
South Africa
Lebanon
Uganda
Zambia
Nigeria
Kenya
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
Y*
Y*
Y*
Y*
Y*
Y*
Y*
Y*
Y*
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
Other
Australia
Russia
Turkey
Y
Y
Y
Y
Y
–
Y**
Y**
–
–
Y
–
–
Y
–
–
Y**
–
Survey coverage continued:
In addition to the manufacturing and service sector PMI surveys, IHS Markit also conducts surveys for construction (UK, Germany, France, Italy and Republic of Ireland), retail (Germany, France and Italy) and jobs (UK) in a number of nations.
An introduction to the PMI surveys
| 11
Coverage: Survey Indices The list of indices available for the manufacturing and service sector surveys is shown below. Reading across the columns of each row of the table shows the comparable index types for each sector and which indices for manufacturing and services are weighted together to calculate the composite readings.
For example, the Manufacturing Output Index is the direct equivalent of the Services Business Activity Index and these two indices are weighted together to calculate the Composite Output Index. (Return to section Data Collection for more information on the weighting process).
The headline index from each survey type is shown in bold text.
Important clarification: The headline Composite Output Index is a weighted average of the headline Services Business Activity Index and the Manufacturing Output Index. The headline Manufacturing PMI is NOT used in the calculation of the Composite Output Index.
Manufacturing Services Composite (manufacturing and services)
Output Business Activity Output
New Orders New Business New Orders
Employment Employment Employment
Input Prices Input Costs Input prices
Output Prices Output Prices Output Prices
Backlogs of Work Outstanding Business Backlogs of Work
Future Activity Future Activity Future Activity
Suppliers’ Delivery Times
Quantity of Purchases
Stocks of Purchases
Stocks of Finished Goods
New Export Orders
Purchasing Managers’ Index*
An introduction to the PMI surveys
| 12
The table below outlines the index coverage for the whole economy, construction and retail surveys:
Whole Economy Construction Retail
Output Total Industry Activity Actual Sales (vs. one month ago)
New Orders New Orders
Employment Employment Employment
Overall Input Prices Input Prices Average Prices Paid
Output Prices
Backlogs of Work
New Export Orders
Quantity of Purchases Quantity of Purchases Value of Goods For Resale
Stocks of Purchases Stocks of Goods For Resale
Suppliers’ Delivery Times Suppliers’ Delivery Times
Purchase Prices
Staff Costs
Future Activity
Purchasing Managers’ Index*
Housing Activity
Commercial Activity
Civil Engineering Activity
Sub-contractor Use
Sub-contractor Availability
Sub-contractor Quality
Sub-contractor Rates Charged
Actual Sales (vs. one year ago)
Actual Sales vs. Plans
Expected Sales vs. Plans
Gross Margins
An introduction to the PMI surveys
| 13
Coverage: Survey Indices – Headline Indices For each survey type (manufacturing, services, whole economy, construction and retail), IHS Markit have defined a concept called the headline index. A headline index is the seasonally adjusted index within that survey which is most closely watched by the markets. The value of the headline index,
unlike the values of other sub-indices, will appear on the press releases published by IHS Markit.
Survey Type Headline Index
Manufacturing
Purchasing Managers’ Index™ (PMI™)
Services
Business Activity Index
Whole Economy Purchasing Managers’ Index™ (PMI™)
Construction
Total Industry Activity Index
Retail Actual Sales vs. One Month Ago Index
For services, construction and retail, the headline index is therefore based on a single question asking if levels activity (or, in the case of retail, sales) are higher, the same or lower than in the previous month.
For manufacturing and whole economy surveys, a separate derived index is assigned headline status. This index is called the Purchasing Managers’ Index™ (PMI™) and is calculated as a weighted average of the following sub-indices:
Survey Type Headline Index
New Orders 30% weight applied
Output
25% weight applied
Employment
20% weight applied
Suppliers’ Delivery Times*
15% weight applied
Stocks of Purchases 10% weight applied
An introduction to the PMI surveys
| 14
* The Suppliers’ Delivery Times Index (SDTi) is inverted, i.e. 100-SDTi, when included in the headline PMI
Coverage: Survey Indices – Flash PMI Indices For certain nations and sectors, IHS Markit also produce a “flash” estimate approximately one week
before final results are published. The flash estimate is calculated in exactly the same way as the final
published data, except the sample is an early cut of the data comprising only 80%-85% of the final
sample.